On Nov 11, 2010, at 10:53 AM, Argyrios Kyrtzidis wrote:

>> 
>> 
>> A more general concern I have about this change (pointed out by this) is 
>> that you're adding translation unit state to the Diagnostic object. I guess 
>> we already had that due to mutation before so this isn't any worse, it just 
>> seems weird to me :)
> 
> That is a good point! How about moving that stuff out into a new 
> DiagnosticState object that is owned by Preprocessor and gets passed to 
> Diagnostic which will have no translation unit state anymore ?
> DiagnosticState will have a SourceManager& and accept SourceLocations, not 
> FullSourceLocs.

Sounds great to me!

-Chris
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to